10/1/2025
Hello Developer Community!!!
My name is Rebecca Pehler. I am a Risk Engineer and need to discuss with the Developer Community a significant ML Compliance potential issue: defining, establishing, and documenting Compliance and Risk Management in response to the efficacy and compliance of using copyrighted material in training and defining performance standards for generative AI.
THE GOOGLE DEVELOPER COMMUNITY: Potential Discussion topic:
Identification of Potential AI Design Risk to Compliance, Change Management, AI Operations start-up, AI Operations reliability, and maintainability. As referred to in this Library of Congress web page, Generative Artificial Intelligence and Copyright Law (https://www.congress.gov/crs-product/LSB10922), this includes but is not limited to the following:
· Do copyrighted works require a human author, and how does this impact the potential copyright application in works created with generative AI?
· May humans copyright works that they create when using AI?
· Does the AI Training process infringe copyrights in other works?
GOOGLE’s Gemini’s Response to define the scope of this potential discussion topic:
Copyright issues with AI center on two main areas: potential infringement by AI models trained on copyrighted material and questions of who owns the copyright to AI-generated content. Lawsuits have been filed by artists, authors, and media companies arguing that AI training violates their rights by reproducing their works without permission. Conversely, U.S. courts have affirmed that works solely created by AI cannot be copyrighted because copyright requires human authorship, but works with significant human creative input that use AI as a tool may be copyrightable.
Library of Congress References:
Generative Artificial Intelligence and Copyright Law (https://www.congress.gov/crs-product/LSB10922)
Copyright Registration Guidance (https://www.congress.gov/crs-product/LSB10922#:~:text=Thus%2C%20the%20AI,register%20their%20copyright.)